The Hard Road to Presentism

The Hard Road to Presentism Jamin Asay Lingnan University Sam Baron University of Sydney Forthcoming in Pacific Philosophical Quarterly Abstract It is a common criticism of presentism the view according to which only the present exists that it errs against truthmaker theory. Recent attempts to resolve the truthmaker objection against presentism proceed by restricting truthmaker maximalism (the view that all truths have truthmakers), maintaining that propositions concerning the past are not made true by anything, but are true nonetheless. Support for this view is typically garnered from the case for negative existential propositions, which some philosophers contend are exceptions to truthmaker maximalism. In this paper, we argue that a no truthmakers approach to the truthmaker objection is critically flawed. 1. Introduction According to presentism, all and only present entities exist. Presentism is opposed to eternalism, according to which past, present and future entities all exist. According to its defenders, presentism, unlike eternalism, captures our everyday intuitions about time. Indeed, this is taken to be one of the central motivating forces in favour of the view (Bigelow 1996, Kierland and Monton 2007, Markosian 2004, Merricks 2007, Tallant 2009b, Zimmerman 2008). However, as Jonathan Tallant (2009b) has recently pointed out, the claim that all and only present entities exist on its own does not do justice to those intuitions. For although one might commonly think that all and only present entities exist, one is also likely to believe that past entities existed and that future entities will exist. Thus, although we don t think that dinosaurs exist they are not out there in the universe some-when, wandering around, getting hit by meteors we do believe that dinosaurs nevertheless existed. Furthermore, we think that only certain things 1

existed: although dinosaurs existed, vampires did not. Intuitively, there is a particular way that the past was. Given the importance that the presentist places on accommodating our intuitions about time, she has good reason to accommodate this further intuition about how things were and will be. But this seemingly innocuous intuition poses a substantial difficulty for presentism. The problem is that, in order to be able to accept the proposition <dinosaurs existed>, say, one must accept that <dinosaurs existed> is true. But according to truthmaker maximalism an important part of standard truthmaker theory for any true proposition P, there exists at least one entity E that makes P true, an entity in virtue of which P is true. (Armstrong 1997, Armstrong 2004). 1 Consider for a moment the claim that Kripke exists. Kripke himself is the obvious candidate to answer the question What makes <Kripke exists> true? But what makes <dinosaurs existed> true? It can t be dinosaurs, since dinosaurs are nowhere to be found in the presentist s ontology. But what then does the relevant truthmaking work? Call this the truthmaking question. As it stands, the truthmaking question is not yet an objection to presentism. In order to develop the truthmaking question into an objection, one requires the further claim that there is nothing available to the presentist that is capable of doing the relevant truthmaking work. With this claim in hand one can formulate a simple truthmaker objection against presentism as follows: 2 The Simple Truthmaker Objection against Presentism (P1) Truths about the past possess truthmakers. [from Truthmaker Maximalism] (P2) If presentism is true, then truths about the past lack truthmakers. [Assumption] (C1) Therefore, presentism is false. [From P1 & P2] There are two ways to resist the simple objection. First, one might reject P1 by arguing that claims about the past simply do not possess truthmakers. If one takes this line, one must reject 1 Truthmaker maximalism is also defended by Cameron (2008) and by Rodriguez-Pereyra (2006) (who defends the view against Milne (2005)). The nature of the truthmaking relation between E and P is controversial; we remain officially neutral on the matter. 2 The simple truthmaker objection focuses on claims about the past. We follow most parties to the debate in thinking that the past-directed form of the objection is the most pressing, but all of our comments apply equally to truths concerning the future. 2

the standard truthmaker theory offered above: the view according to which, for every truth there is at least one entity that makes that truth true. There are two options here, one radical and one modest. First, one might reject truthmaker theory tout court (and not just maximalism). On this view, there simply are no cases of truthmaking, not even between Kripke and <Kripke exists>. The connection between truth and reality is thereby severed. We think, however, that there are two good reasons not to proceed in this fashion. First, truthmaking provides a needed connection between true beliefs and ontology. If a belief such as the belief that there is a mind-independent external world is true, then there must be some mindindependent thing that makes this belief true, namely, the external world. If, however, truthmaker theory is false, and true beliefs need not depend on what exists, then the belief that there is a mind-independent external world can be true, without there being anything whatsoever that makes this claim true. So for all we know, the external world does not exist, even though we believe truly that it does. Hence, by giving up truthmaker theory, one courts skepticism. Second, rejecting truthmaker theory tout court suggests that what s true doesn t depend on reality at all. But this seems to pose a problem even in non-skeptical cases. For example, if <Kripke exists> doesn t depend on what exists for its truth, then its truth seems to be independent of the existence or non-existence of Kripke himself. We think, however, that it is obvious that the truth of <Kripke exists> depends on Kripke. 3 Indeed, that the truth of <Kripke exists> depends on Kripke possesses Moorean certainty: it is something that we know better than any philosopher s argument to the contrary. Hence, the obviousness of this dependence performs an immediate reductio on any view that treats the truth of <Kripke exists> as independent of Kripke s existence. The second and more modest option for rejecting P1 is to reject truthmaker maximalism, but without abandoning truthmaker theory altogether. Although most truths require truthmakers, truths about the past do not. While this option does not seem to pose any skeptical problems, this solution to the simple objection constitutes the hard road to presentism. The road is hard because, according to truthmaker theorists like Ross Cameron (2008), there is no plausible way to restrict truthmaker theory without rejecting truthmaker theory altogether: truthmaking is an all-or-nothing affair. So the presentist who takes this option has work to do: she must show that 3 When we say that the truth of <Kripke exists> depends on Kripke, we are not expressing the truism that <Kripke exists> iff Kripke exists, which is an instance of the Tarskian truth-schema. Nor are we saying <Kripke exists> because Kripke exists, if the because connective is understood to relate propositions. Rather we are saying that there is some ontological dependence relation between the truth of <Kripke exists> and the world. It is the obviousness of that dependence relation that, we think, militates against rejecting truthmaker theory tout court. 3

truthmaker theory is not all-or-nothing by motivating and defending a restriction of the relevant kind. The second response to the simple truthmaker objection is to reject P2: even if presentism is true, it does not follow that claims about the past lack truthmakers. Indeed, according to this view, perfectly good truthmakers are on hand to make statements concerning the past true. For example, John Bigelow (1996) argues that there are presently instantiated, tensed worldproperties that can do this work; the property having once possessed dinosaurs, for instance, makes true the proposition <dinosaurs existed>. Alternatively, consider Tom Crisp s (2007) ersatzer presentism (see also Bourne (2006a)). According to ersatzer presentism, the truthmakers for claims about the past are abstract entities: maximally consistent sets of tenseless propositions ordered by the E-relation - a successor relation over abstract objects, analogous to the earlierthan relation that relates concrete times under eternalism. This second solution to the simple truthmaker objection is the easy road. The road is easy because it does not require restricting truthmaker theory. Rather, it takes truthmaker theory in its stride and accepts the challenge to provide truthmakers for claims about the past. The easy road is the road well-travelled (see, for example, Cameron (2011), Rhoda (2009) and Keller (2004)). Lately, however, the easy road has begun to look treacherous. In a recent series of papers David Sanson and Ben Caplan (2010, Caplan and Sanson 2011) argue that all extant forms of the easy road are explanatorily deficient. Tensed properties, abstract entities and so on do not adequately explain why it is that truths about the past are true. And, as David Liggins (2005) and Ian McFetridge (1990) have argued, there is good reason to suppose that truthmaking is (at least) constrained by explanation: what it is to make something true is, at least in part, to explain why it s true. The Sanson and Caplan line effectively shuts down the easy road: on their view, the only entities capable of explaining why truths about the past are true are past entities. Hence, no form of the easy road can discharge the relevant explanatory burden, since no form of the easy road deploys past entities as the truthmakers for claims about the past. 4 The Sanson and Caplan critique is likely to push presentists down the hard road. Our goal, then, is to head presentists off at the pass by showing that hard road strategies fail to furnish the presentist with a solution to the truthmaker objection. It should be noted, however, that we are not concerned to show that it is impossible, in general, to motivate a restriction on truthmaker theory. Although we are sympathetic to this line of thought, establishing this claim would take us too far afield. Rather, our goal is the more modest one of showing that one cannot motivate the kind of restriction appealed to by hard road presentists to get presentism out of strife. In 4 Merricks (2007) and Baron (2012) use similar considerations to place pressure on easy road strategies. 4

pursuing this goal we will focus primarily on the hard road strategy of Jonathan Tallant (2009b, 2009a), as it reflects the state of the art in hard road strategies, and because any form of the hard road is likely to proceed along similar lines. 2. Milestones In his (2009b), Tallant employs a partners in crime defense of presentist truthmaking by turning to the notorious case of negative existential truths, such as <there are no unicorns>. Two dominant strategies for handling negative existentials are: (A) positing entities such as negative facts to serve as truthmakers (see Russell (1919) and Armstrong (2004)) and (B) maintaining that negative existentials do not have truthmakers (see Bigelow (1988) and Lewis (2001, 1992)). Tallant suggests that presentists can avail themselves of either strategy. Presentists can posit the existence of tensed facts that are no more worrisome than negative facts. Or they can maintain that truths about the non-present are just as good candidates for being truthmaker gaps (i.e., truths without truthmakers) as are negative existentials. The first of Tallant s two options here is a form of the easy road, since it accepts the need to supply truthmakers for truths about the past, and so proceeds by capitulating to a suitably maximalist truthmaker theory. Although interesting, we will set Tallant s defense of the easy road aside, for it appears to be undermined by Sanson and Caplan s aforementioned concerns regarding explanation. 5 The second option is a hard road strategy because it places a twofold restriction on truthmaker theory, with regard to both negative existentials and truths about the past. 6 There are thus three milestones along Tallant s hard road. First, there is the claim that negative existentials lack truthmakers, which sets a principled precedent for restricting truthmaker theory. Second, 5 There are good reasons not to follow Tallant down his easy road. First, tensed facts do not seem to be good partners in crime with negative facts. For one thing, Armstrong only needs his one totality fact. A presentist who endorses tensed facts, however, will require an indefinite amount. Second, there are substantive questions facing presentism concerning the nature of tensed facts. What are tensed facts composed of? Because of the presentist s thrifty ontology, tensed facts must be made up of either present entities or abstract entities. This will lead toward something like Bigelow s tensed properties view or to Crisp s ersatzer presentism. Thus, Tallant s easy road will ultimately collapse into the standard easy road to presentism, which has been previously criticised by Sanson and Caplan. 6 In principle, one need not develop the hard road via negative truths in this manner. One might argue for a restriction on claims about the past only, allowing that negative truths have truthmakers. Although one could take this line, no one in the literature has, and it is hard to see how one might motivate a restriction for claims about the past other than via an analogy to negative truths. 5

there is the claim that, because we can restrict in the case of negative existentials, we can restrict in the case of claims about the past as well. And third, Tallant offers new, restricted, truthmaking principles, which are supposed to accommodate truthmaker gaps with regard to both negative existentials and the past. Challenging each of these three claims provides an opportunity to block the hard road. In what follows, we consider each milestone in turn, and focus our criticisms on the latter two. 3. Step One: Negative Existentials Why think that negative existentials lack truthmakers? Well, on standard truthmaker theory, when some object is a truthmaker for some proposition, the proposition is true in virtue of the existence of that object. <Penguins exist> is true in virtue of the existence of penguins. But a negative existential is a claim about what doesn t exist. On the face of it, it s true not in virtue of the existence of something of a particular kind, but in virtue of the fact that nothing of a different kind exists. In effect, the truthmaker maximalist demands that an atheist about, say, unicorns, must ground her atheism with a theism about something very different, such as an absence (Martin 1996), negative fact (Russell 1919), or totality state of affairs (Armstrong 2004). What s hard to appreciate is how one might motivate the positing of absences, negative facts, or totality states of affairs independently of one s theoretical commitment to maximalist truthmaker theory. Now, there might be good overall reasons for accepting maximalist truthmaker theory, biting the bullet, and countenancing objects like totality states of affairs. (This is the perspective that Armstrong (2004) takes.) In that case, the hard road is shut down early: if non-maximalist truthmaker theory is indefensible, then so too is the hard road to presentism. But the important point to notice is that the very idea of truthmaking that existing objects make propositions true might be seen as misplaced when applied to negative existentials, which are about what doesn t exist. The hard road presentist needs to argue similarly: if we can motivate the idea that some truths, like negative existentials, lack truthmakers, then we begin to motivate the idea that truths about the past also lack truthmakers. In a recent paper, however, Cheyne and Pigden (2006) offer a straightforward position according to which negative existentials do, in fact, possess truthmakers and, moreover, that all the truthmaking work can be done with things that are already on hand. Thus, they contest the (albeit intuitive) idea that truthmaking is misplaced when it is applied to negative existentials. If their argument is successful, the hard road presentist will fail to reach the first milestone. To gain a sense of the Cheyne and Pigden line, consider the following proposition: 6

(1) <there is no hippopotamus in room S223> In order to make (1) true, argue Cheyne and Pigden, we require only a single, positive, truthmaker: the way that room S223 actually is on 01/02/12, since the way that room S223 actually is on 01/02/12 (allegedly) guarantees the absence of a hippo in the room in question, and so the way that the room S223 actually is on 01/02/12 serves to guarantee the truth of (1). Cheyne and Pigden subsequently extend this idea to more general negative existentials, like: (2) <there are no unicorns> They contend that the correct truthmaker for (2) is the (first-order) way the universe actually is which is a large, complex fact, but a positive fact for all that. As with the actual way that room S223 is, the way the actual world is (again, allegedly) guarantees that there are no unicorns and thus guarantees the truth of (2). (A similar view has been offered by Cameron (2008) and Schaffer (2010).) The advantage of the Cheyne and Pigden view is that, if it s successful, we need no new ontology to make negative existentials true: no absences, totality facts or negative facts need apply. Rather, all the truthmaking is fixed by the positive way that the world actually is: a truthmaker which, presumably, we already have available. But more than this, it shows that truthmaking is not misplaced when it comes to negative existentials: if these facts guarantee the truths in question, then negative existentials are appropriate candidates for truthmaking after all no restrictions are necessary. If Cheyne and Pigden are correct, then the first leg of the hard road to presentism is a dead end: we no longer have any reason to think that negative existential propositions lack truthmakers, and so no way to use negative existentials at the second leg to make a case for restricting truthmaker theory in the case of presentism. But, fortunately for the hard road presentist, there are two good reasons to doubt the complete success of the Cheyne and Pigden line. First, in a recent paper, Parsons (2006) identifies a serious difficulty for the idea that the world serves as the truthmaker for negative existentials like <there are no unicorns>. Parsons begins by differentiating between two different kinds of negative claims: partially general negative claims and fully general negative claims. Partially general negative claims, such as (1) above, posit the non-existence of a kind of entity within a limited region of space and/or time. Fully general negative claims are not restricted in this fashion, as in a claim like (2). This claim is fully general 7

in that it says that there are no unicorns simpliciter, not merely that there are no unicorns in this or that part of the universe. Parsons concedes that the Cheyne and Pigden strategy succeeds for partially general negative claims: once we fix the actual way that room S223 is at a particular time, we fix the fact that there is no hippo in that room at that time and so the actual way that room S223 is guarantees the truth of that claim. Similarly, Parsons admits that what we might call specific negative predications like <Theaetetus is not flying> are amenable to the Cheyne and Pigden analysis; these are claims that apply a negative feature to a specific individual. Theaetetus himself, sitting as he is on the ground, guarantees the truth of <Theaetetus is not flying>. However, according to Parsons, the Cheyne and Pigden line fails for fully general negative claims like <there are no unicorns>. To illustrate this, Parsons offers the following example. Suppose that we take the actual universe and situate it within a broader reality in which it is but one of two island universes. Suppose that in our universe, universe A, unicorns do not exist, whilst in universe B they do. The (first-order) way that universe A actually is whatever it is, in other words, that Cheyne and Pigden believe makes <there are no unicorns> true still exists in the A-B universe. But <there are no unicorns> is false in the A-B universe. So Cheyne and Pigden s offered truthmaker fails to guarantee the truth of <there are no unicorns>; they have yet to find truthmakers for fully general negative existentials. This brings us to the second reason to doubt the complete success of the Cheyne and Pigden line. Consider the partially general truth (1). According to Cheyne and Pigden, it is the actual way that room S223 is on 01/02/12 that guarantees the truth of that claim. As Armstrong (2006) argues, however, it is hard to see how this contingent, positive fact about room S223 might genuinely guarantee that there is no hippo in room S223. This becomes more evident when we substitute the actual way that room S223 is on 01/02/12 with the phrase a certain collection of first-order positive states of affairs. Put this way it is, in Armstrong s words, opaque as to why the actual way that room S223 is on 01/02/12 should guarantee the truth of (1). It is worth pointing out, however, that while Armstrong s worry appears intuitively compelling for the cases of partially general negative truths (about which, it will be recalled, Parsons was relatively sanguine), it does not seem compelling for at least some specific negative truths, like <Theaetatus is not flying>. In this case, it does seem that Theaetatus sitting on the ground guarantees the truth of <Theaetatus is not flying>: every world in which Theaetatus is sitting on the ground is a world in which Theaetatus is not flying (cf. Veber (2008)). 7 7 As an anonymous referee has pointed out to us, there is a further problem for the Cheyne and Pigden line. The actual way that room S223 is isn t compossible with its containing a hippo, and that s why the actual way that room 8

We think that these responses to Cheyne and Pigden are compelling (if not decisive) and, at the very least, provide a basis upon which to develop the first step along the hard road to presentism. But although Cheyne and Pigden s strategy might fail for all negative truths, we think that the dialectic here is instructive, for the following reason. Parsons shows us that the Cheyne and Pigden line fails for fully general negative claims and Armstrong shows us that it fails for partially general negative claims. However, the Cheyne and Pigden line does enjoy some limited success when it comes to certain specific negative claims, such as negative predications. This will become important in what follows. For now it suffices to say that in order to dodge the objections from Cheyne and Pigden that challenge non-maximalist truthmaker theory, hard road presentists may avail themselves of the criticisms offered by Parsons and Armstrong. 8 Of course, should these criticisms fail (and Cheyne and Pigden, or some other maximalist strategy, succeed), then hard road presentism is shut down. In what follows, we shall assume that this is not the case, and thus that we may continue exploring the hard road. 4. Step Two: The Analogy Suppose one manages to motivate a restriction on truthmaking via negative existentials. In so doing, one must be wary of explosion: a situation where one s restriction on truthmaking motivates truthmaking restrictions on all propositions, thereby undercutting truthmaker theory entirely. This would be problematic: as already discussed, no one should give up truthmaker theory tout court. If one is concerned to avoid explosive restrictions one cannot argue simply that because truthmaker theory can be restricted in the case of negative existentials, any restriction whatsoever is therefore acceptable. Rather, if one wishes to use the negative existential case to extend the restriction to a further domain of discourse, one must show that the claims in that discourse are analogous to negative existential propositions in important respects, and that is why S223 is makes true <there are no hippos in room S223>. However, in order to spell this out, one must appeal to certain universal facts about room S223 such as <room S223 contains only air>. Since these universal facts are equivalent to negative existentials, the truthmaker for <there are no hippos in room S223> is not entirely positive. As with Armstrong s worry, however, this difficulty does not seem to generalise to <Thaeatatus is not flying>. The truthmaker for this claim namely, Thaetatus can be spelled out without the need for universal propositions. In every world in which Thaeatatus is sitting on the ground, he is not flying, so his sitting on the ground (a positive fact) is not compossible with his flying. 8 Although Armstrong criticises the Cheyne and Pigden line, he does not conclude against truthmaker maximalism, because he endorses the existence of totality facts. Nevertheless, the presentist might avail herself of Armstrong s criticism of Cheyne and Pigden s view without taking this further step, thereby opening up the hard road. 9

they need not possess truthmakers. For if they need not be analogous then it is hard to see how the path from negative existentials to the claims in the other discourse might be anything other than explosive. This applies to hard road presentism: in order to motivate a restriction on truthmaking in the case of truths about the past, such truths must be appropriately analogous to negative existentials. But presentists like Tallant have not yet told us in what respects claims about the past are analogous to negative existentials. What Tallant does instead is state what a non-maximalist presentist should say, and then attempts to explain away the felt need for truthmaking when it comes to propositions about the non-present. Tallant writes: If we look to the world, claims the presentist, we will find no existent entities to act as truthmakers for talk about what was. Thus, because there are no existent entities that truth-make past-tensed propositions so there can be no demand for truth-makers for past-tensed propositions (2009b: 414). This response, of course, is entirely question begging when offered to the eternalist. What is needed is some argument, independent of the debate between presentists and eternalists, for why truths about the non-present are appropriately analogous to negative existentials and thus do not require truthmakers. We offer three considerations that cast doubt on the availability of an argument of this kind. 4.1 Logical Variety Recall our discussion of negative existentials in 3. As noted, there is good reason to suppose that partially general negative claims and fully general negative claims lack truthmakers. The presentist who opts for the hard road thus needs to establish an analogy between negative claims of this kind and truths about the past. However, many of the truths about the past that presentists are concerned to preserve are not partially general or fully general negative claims. In fact, presentists uphold the truth of all sorts of claims that run the gamut from general to specific, both positive and negative. For instance, all of the following are true: 10

<there were dinosaurs> [fully general, positive] <there have been no unicorns> [fully general, negative] 9 <there were dinosaurs on Earth> [partially general, positive] <there were no dinosaurs on Mars> [partially general, negative] <Socrates was a philosopher> [specific, positive] <Socrates was not a sophist> [specific, negative] The trouble is that hard road presentists need to establish that all of the truths about the past that are intuitively true are also relevantly similar to fully/partially general negative claims, if they are to defend the view that truths about the past systematically do not require truthmakers. But the truths about the past do not conform to one single logical form, and so the analogy is flawed. Indeed, it is flawed in two senses. First, many of the claims about the past that are intuitively true are disanalogous to fully general/partially general negative claims in terms of their scope. We have in mind here propositions like <Socrates was not a sophist>. If anything, this claim is analogous to specific negative predications like <Theaetatus is not flying>. But as discussed in 3, claims like <Theaetatus is not flying> have truthmakers. The truth of these claims is guaranteed by certain positive features of the world, such as Theaetatus sitting. So there is no way to move from a restriction on truthmaking for negative claims of this kind to claims with an analogous logical form about the past. Second, many claims about the past are not negative claims. When we say that <Socrates was a philosopher> is true, or <there were dinosaurs> is true we are saying something positive about the past. Hence, if anything, these claims are analogous to ordinary positive claims that possess truthmakers. Granted, these claims are analogous to negative claims in terms of their scope: <there were dinosaurs> is, for the presentist at least (see fn. 7), fully general, just like <there are no unicorns>. But it is because of the negativity of negative claims that a restriction to 9 It could be argued that <there were no dinosaurs> and <there have been no unicorns> are fully general with regard to space, but are only partially general with regard to time because they are restricted to the past. However, we take it that for the presentist these are fully general with respect to time, as are all claims, since there is only one moment of time if presentism is true. If, however, we are wrong and <there were no dinosaurs> is not fully general, then what we say below regarding the scope of that claim should be modified to say that <there were no dinosaurs> is analogous to <there are no unicorns on Earth>, in that it is a partially general claim. Even if this is correct, our point still stands: there is a disanalogy between positive claims about the past and negative claims about, say, unicorns. 11

maximalism is motivated, and so an analogy in scope is not enough on its own to motivate a truthmaker restriction for propositions about the past. In sum: the hard road presentist needs to establish that some negative existentials are plausibly taken to be truthmaker gaps, and thus may rely on either Armstrong s 10 or Parsons criticisms of Cheyne and Pigden s defense of maximalism. But the space for non-maximalist truthmaker theory that these criticisms open up provides no helpful analogy for the presentist: the kinds of negative truths that plausibly lack truthmakers are not analogous to the wide variety of truths about the past. 4.2 In the Fiction This brings us to the second reason for doubting the analogy between truths about the past and negative existentials. Consider a range of fictional propositions about the past, such as the proposition: (3) <vampires existed>. Now, compare (3) with (4): (4) <dinosaurs existed>. Presumably, (4) is true whilst (3) is false. Indeed, more generally, there must be a distinction between fictional claims about the past and non-fictional claims about the past along these lines. For if this distinction is not viable, and (3) and (4) share the same truth value, then there is no way to distinguish between the genuine history of the universe and some merely fictional history. But this is a distinction that one clearly can and should be able to make. For if one cannot make this distinction, then one has no basis for saying that although dinosaurs existed, vampires did not, and so no way to accommodate the intuition that there is a particular way that the past was, which is one of the central intuitions that all presentists must accommodate. But on what grounds might the presentist maintain that (4) is true whilst (3) is false? One obvious option is to hold that only (4) possesses a truthmaker, while (3) does not. This, however, is not a claim that the hard road presentist can hope to defend, since for them (4) lacks a truthmaker and is true nonetheless. Thus, if lacking a truthmaker were a reason to think that a 10 As above: although Armstrong criticizes Cheyne and Pigden s view, he retains his commitment to maximalism. 12

proposition about the past is false, then that would cut just as badly against the truth of (4). Hence, for the hard roader some further reason is required for thinking that only (4) is true. Now, one might think that the matter is relatively straightforward: the distinction between (3) and (4) should be drawn where it appears to be, in the difference between the past and the fictional. 11 But what is the difference between the past and the fictional? A reasonable proposal might be the following: past entities once possessed some positive ontological status, whereas fictional entities never possessed a positive ontological status. Thus, (3) is false because there never were any vampires and (4) is true because there really were dinosaurs. This suggestion, however, merely defers the problem; the proposed explanation presupposes the very distinction between the genuine history of the world and some merely fictional history for which we are trying to account. Specifically, dinosaurs once had a positive ontological status, whilst vampires did not, just in case (4*) is true and (3*) is false: (3*) <vampires once had a positive ontological status> (4*) <dinosaurs once had a positive ontological status> But now it would seem the problem recurs: on what grounds might the hard road presentist indict (3*) but not (4*)? It cannot be because (3*) lacks a truthmaker whilst (4*) does not because, again, lacking a truthmaker is no bar to being true when it comes to claims about the past. Moreover, it will not do to re-draw the distinction where it appears to lie, in terms of the distinction between what is past and what is fictional, and then use that distinction to show that (3*) is false whilst (4*) is true, because it is that very distinction that is now at stake. Rather, what the hard road presentist needs is a new way to draw the distinction between the past and the fictional, one that is not open to the same move. Can such a distinction be provided? We are doubtful. The trouble is that any condition one might place on dinosaurs over vampires such that dinosaurs are past whereas vampires are fictional will be a condition along the lines of (4*): a condition that says of a certain class of entities that they were thus and so. So there appears to be no way to draw the needed distinction except by stating the condition in a way that will require further claims about the past. One will then need a further reason to think that only claims of this sort about dinosaurs are true and thus that the condition holds in one case but not the other. Thus, the distinction one requires will always be deferred, and never drawn. 11 Our thanks go to an anonymous referee for this suggestion. 13

Of course, the worry simply does not arise if propositions about the past possess truthmakers. If propositions about the past possess truthmakers, then we can say that (4) is true whilst (3) is false, because (4) has a truthmaker whilst (3) does not, a condition that can be stated without the need for further claims about the past. Thus, it is only if one embraces the analogy between propositions about the past and negative existentials, thereby granting that propositions of the former kind also lack truthmakers, that one loses the distinction between fictional and nonfictional claims about the past. Since we clearly can differentiate between these two kinds of claims there is therefore reason to doubt the viability of the analogy at issue. 4.3 Dialectical Troubles Our third and final objection to the second stage of the hard road concerns the broad dialectic in which the presentist is engaged. Suppose that all negative existentials lack familiar, ontologically non-controversial truthmakers. In such a situation, the debate over negative existentials is between whether we should (i) posit a new kind of entity to satisfy standard truthmaker theory or (ii) argue that truthmaking ambitions are misplaced when it comes to negative existentials. Presentists like Tallant aim to make the case that these two paths are also open with respect to truths about the non-present, and let us grant that they are (supposing that the hard road presentist can meet the objections already offered). But there is a third option when it comes to truths about the past and future: we can (iii) posit familiar kinds of entities to satisfy maximalism. In other words, the eternalist need not face the parallel worry that truthmaker theorists face with respect to negative existentials. There is a happy alternative: what makes claims about the past and future true are exactly the same sorts of entities that make claims about the present true. Kripke makes true <Kripke exists>, and Socrates makes true <Socrates existed>. Whatever makes it true that the irises in the garden are blue is the same kind of thing that makes it true that the irises from last year s garden were blue. Eternalists need not posit controversial tensed facts, nor need they rely on an analogy to negative existentials in order to free themselves from the demands of truthmaker theory. As a result, the truthmaking objection to presentism plays right into the hands of the eternalist. Even if the presentist has two available responses, the best response remains the eternalist one and so presentists who take the hard road have not, as yet, provided a way to deflect the truthmaker objection. Hence, the debate regarding truthmakers for negative truths is not sufficiently analogous to the debate regarding truthmakers for truths about the past. There 14

are dialectical options available in the latter debate that do not arise in the former. As a result, hard road presentists can find no shelter by turning to the debate over negative truths. 5. Step Three: Weak Truthmaker Principles There is therefore cause for doubt concerning the second stage of the hard road to presentism. In this last section we will argue that even if one could make the analogy between negative existentials and claims about the past work, it is far from clear that there is a candidate, weakened truthmaker principle that will serve the needs of the hard road presentist. There appear to be just two such principles on offer. The first principle is offered by Tallant: (NGC-ST) a proposition is true if and only if, either: (a) there exists an entity that makes that proposition true; or, (b) there does not exist an entity and that makes the proposition true; or, (c) there could have existed an entity that would make the proposition true; or, (d) there has existed an entity that makes the proposition true... (2009a, p. 426) (NGC-ST) is a non-starter. Take any contingently true negative existential, such as <there are no Arctic penguins>. Tallant takes such propositions to satisfy condition (b). But its negation, <there are Arctic penguins> satisfies condition (c), and so (NGC-ST) upholds the truth of both a proposition and its negation. (NGC-ST) is dead on arrival. In responding to Krämer (2010), who notices this same problem, Tallant acknowledges that this is a bad result (2010, p. 503) and in its place offers: (CGP) For every proposition, that proposition is true iff it accurately characterizes its subject matter. As a replacement truthmaking principle, (CGP) has little to offer. To say that a proposition accurately characterizes its subject matter is no more than a roundabout way of saying that a proposition is true. Tallant s own examples of the principle bear this out: <there were dinosaurs> accurately characterizes the world, he says, if and only if dinosaurs have existed. ( there were dinosaurs and dinosaurs have existed, it goes without saying, are two ways of saying the same thing.) So (CGP) is, in fact, nothing more than a statement of the propositional truth schema: 15

For every proposition <p>, <p> is true iff p. Hence, (CGP) is not a truthmaking principle at all. It merely points to the same set of equivalences that, according to Tarski (1956), every adequate theory of truth must entail. 12 But these equivalences are neither here nor there when it comes to accounting for the sort of dependence at issue in truthmaker theory. If they were, then everyone would, quite trivially, be a truthmaker maximalist. For if all there is to a proposition s having a truthmaker is there being some sentence materially equivalent to <p> is true, then every proposition (or, at least, every proposition we can name) has a truthmaker. Alternatively, if upholding the truth schema is all there is to capturing the insight behind truthmaking, then there is nothing to motivate the ontological debates that fuel truthmaker theory. Do all truths have truthmakers, or are there truthmaker gaps? Are states of affairs required to serve as truthmakers, or do tropes suffice? What is the nature of the dependence between a truth and its truthmaker(s)? Such questions disappear if everything there is to say about truthmaking can be captured by (CGP). 13 (CGP), then, is not a plausible reformulation of (NGC-ST). Is there a more plausible way to redevelop that principle? One option might be to appeal to the supervenience version of truthmaking defended by John Bigelow (1988) and David Lewis (2001). The supervenience view does not contend that for every truth, there must exist a truthmaker. Rather, this truthmaker principle requires only that the truth of a proposition supervenes on what exists. Following Lewis, we state this truthmaker principle as follows: The Supervenience Principle (SP): For any proposition P and any worlds W and V, if P is true in W but not in V, then either something exists in one of the worlds but not in the other, or else some n-tuple of things stands in some fundamental relation in one of the worlds but not in the other. (Lewis 2001, p. 612). 14 12 Technically, Tarski s requirement involved sentential versions of the schema, not propositional ones: p is true iff p. 13 To push this point further, consider that all parties to the debate agree that <there are no unicorns> is true iff there are no unicorns. But that in no way settles the contentious question about what, if anything, exists that makes it true. 14 As Lewis notes, this formulation of the supervenience principle is weaker than Bigelow s formulation. Bigelow s principle is formulated without the second disjunct. Thus, for Bigelow, all that the supervenience principle requires is that for any proposition P and any worlds W and V, if P is true in W but not in V, then something exists in one of the worlds but not the other. 16

It is commonly agreed that SP avoids the need to provide truthmakers for negative existential propositions. As Lewis writes: [SP] allows truths to have truthmakers, but also it allows them to be true just because they lack falsemakers. The simplest case is that of a negative existential: the proposition that there are no unicorns, say. It is true in the actual world just because there are no unicorns to make it false. In any world where it is false, certain things would have to exist which in actuality do not exist, namely one or more unicorns. (Lewis 2001, p. 610) If Lewis is correct then one might argue as follows: if negative existentials can be true because of how things are not under (SP) without committing to any truthmakers, then perhaps claims about the past can be true because of how things were without committing to any truthmakers. (SP), it could be argued, leaves this entirely open, and so a principle of this kind might be wellpositioned to serve the hard road presentist s need for a principle that eschews truthmakers for claims about the past. Thus, Tallant might fruitfully reformulate (NGC-ST) by combining it with something like (SP), to produce the following tensed account of truthmaking: The Supervenience Principle* (SP*): For any proposition P and any worlds W and V, if P is true in W but not in V, then either something exists, existed or will exist in one of the worlds but not in the other, or else some n-tuple of things stands, stood or will stand in some fundamental relation in one of the worlds but not in the other. The idea behind (SP*) is to widen the supervenience base for propositions. 15 Propositions no longer supervene merely on what exists. Rather, propositions supervene on what exists and what existed (and will exist). Thus propositions can be true in virtue of how things were, which appears to capture condition (d) of (NGC-ST) in a more plausible form. By widening the supervenience base in this fashion, the hard road presentist avoids the need to provide truthmakers for claims about the past: just as negative existential propositions can be true because of the lack of falsemakers, propositions about the past can be true in virtue of what merely existed. There is, however, a serious difficulty with (SP*). According to (SP*), the truth of propositions about the past supervenes on what existed. But what existed, according to the presentist, does not (ordinarily) exist. So propositions about the past supervene on what does not exist. Supervenience is, however, a relation: if a supervenes on b then that is in virtue of some 15 Again, our thanks go to an anonymous referee for this suggestion. 17

supervenience relation R such that arb. Consequently, if the truth of propositions concerning the past supervenes on what existed (but does not exist), then there must exist cross-temporal supervenience relations that have as their relata entities that merely existed (but do not exist). However, most presentists deny the existence of relations of this kind because most presentists are serious presentists, where serious presentism is the view according to which past entities entities that existed do not possess properties or stand in relations (Bergmann 1999, pp. 125 126, Markosian 2004, p. 52, Hinchliff 1996, pp. 124 126 ). This form of presentism (at least with regard to relations) is motivated by a commitment to the intuitive idea that all relations are existence entailing (which Crisp (2005, p. 5) claims is a truism ), where a relation N is existence entailing iff necessarily, xny only if x and y exist (Bigelow 1996, p. 37). Commitment to (SP*) therefore requires rejecting existence entailment and, with it, serious presentism. It is commonly thought, however, that the rejection of serious presentism leads to an unattractive Meinongian metaphysics (see, for instance, Brogaard (2006, p. 195), Bergmann (1999) and Markosian (2004, p. 52)), according to which there are past things that instantiate fundamental properties and serve as the relata in cross-temporal relations, though they do not exist. Presentists, however, are not generally willing to adopt Meinongianism with regard to the past for the simple reason that once one has bought into Meinongianism with regard to nonpresent things, there no longer appears to be any clear distinction between presentism and nonpresentism to be drawn (Keller 2004, pp. 89 91, Lewis 2004a, pp. 7 11), and certainly no clear motivation for endorsing one view over the other. (Nor is it clear why the Meinongian strategy would not simply collapse into the easy road, for it appears that Meinongian objects could serve as truthmakers for claims about the past.) Now, one might disagree with this objection against (SP*) on the following grounds: if the supervenience needed for (SP*) requires relations, then the supervenience needed for (SP) requires relations also. But this means contra Lewis that (SP) cannot be used to handle the truth of negative existential claims. For a similar problem can be raised there: the truth of <there are no unicorns> under (SP) supervenes on what is not rather than what exists. But supervenience requires relations and all relations are existence entailing. So it seems we must be committed to the existence of things that are not. Such things, however, can only be absences of some kind. But this result flatly contradicts what Lewis says, namely that (SP) enables one to accommodate the truth of negative existential claims without committing to the existence of absences. However, it is somewhat misleading to say that, given (SP), negative existentials supervene on what is not, at least if this is taken to mean that negative existential do not supervene on what exists. It is clear, rather, that for Lewis negative existentials that obey (SP) supervene on what 18

exists; what it means for a negative existential proposition P to obey (SP) is just that if P is true at a world W and false at a world V, then there is something that exists at W and not at V or vice versa. So, for example, consider <there are no unicorns>. This proposition obeys (SP) in the sense that if that proposition is true at a world W and false at a world V, then unicorns exist at V but not at W. It is not the case, however, that the truth of <there are no unicorns> thereby supervenes on what is not and thus requires the reification of the absence of unicorns. Rather, the truth of that proposition supervenes on the ontology of the world, in the sense that if you make a difference to the truth of <there are no unicorns> then you must make a difference to the ontology. The supervenience relation is always connecting things that exist, even in the case of negative existentials. These considerations do not extend to (SP*): it is not the case that propositions about the past supervene on the ontology of the world. To see this, consider the proposition <dinosaurs existed>. Suppose that this proposition is true at a presentist world W and false at a presentist world V. On the face of it, this difference in truth value makes no difference to the ontology of the worlds: a presentist world in which <dinosaurs existed> is true and a world in which <dinosaurs existed> is false might nevertheless be the same in all relevant ontological respects. Indeed, if a change to the truth-value of <dinosaurs existed> between W and V did make a difference to the ontology of the two worlds, then some account would be required of exactly what that difference might be. Now, it cannot be a difference in dinosaurs, because dinosaurs (we can assume) do not exist in either world. So there must be some further difference in the ontology that accounts for the difference in truth. The challenge for the presentist who accepts (SP*) and who wishes to retain serious presentism is to spell out exactly what this difference might be. The presentist of this kind must now add something to the ontology that can serve as the supervenience base for the truth of <dinosaurs existed>. However, this has all the hallmarks and troubles of the easy road to presentism. Thus, assuming that the hard road presentist wants to stay on the hard road at this juncture, she must accept that propositions about the past, unlike negative existentials, do not supervene on what exists but, rather, supervene on what existed which, as we have argued, requires relations to non-existent entities. (SP*) then does not appear capable of serving the hard road presentist s needs: if propositions about the past satisfy (SP*), then that is either because one thinks that non-existent entities can serve as the relata in cross-time relations or because one is presupposing some form of the easy road to presentism. Neither option looks attractive. The failure of (SP*) is, however, instructive. (SP*) fails, it would seem, for the following reason: although it avoids the need to provide 19